A few days ago, I posted an article on RepRap. The purpose of the article was to use RepRap as an example of how to think about the economic impact of manufacturing systems that could duplicate themselves, or parts of themselves. In that article, I concluded that RepRap's current version probably wouldn't have much impact.
I'm writing this article to say that I was at least half wrong.
Since I wrote that last article, RepRap has been used to make a circuit board that has been used in the RepRap machine.
This puts RepRap far ahead of most (all?) other rapid prototyping systems in terms of the range of products it can make. It's already comparable on speed. And RepRappers are already thinking about making 3D circuits with embedded components.
This circuit board took 3 minutes to deposit the solder. That's quite a lot of value per hour. The current circuit is a bit "lumpy" but I expect that tuning the process parameters will fix that pretty quickly.
If RepRap continues improving at this pace, it could go from marginally significant to a major competitor in under a decade.
A quick google found several interesting articles. This article on additive manufacturing trends says that we'll be seeing commercial machines under $5,000 soon, and that they will mainly be a vehicle for selling expensive consumables. RepRap's consumables are very inexpensive, and it is cheaper than even the lowest-end commercial product. So if there's an economic case for buying one of those, there's a stronger economic case for RepRap (assuming comparable quality - which seems pretty safe to assume).
This article on plastic capabilities and materials shows that there's both demand for, and availability of, material properties for a variety of demanding applications. Additive fabricated parts are now used in airplanes and bulldozers.
This article on the market for additive fabrication says that it's already crossed the $1 billion mark and is expected to reach $3.5 billion by 2015.
I wonder how long it'll be before we see a for-profit company making packaged, commercialized, user-friendly machines based on RepRap technology, the way Red Hat does for Linux? I wonder how long it'll be before that company makes its machines on a room full of RepRap's... or perhaps it will pay local hobbyists to produce parts when their machines are otherwise idle?